Chemistry & Dynamics of the Milky Way From Before Hipparcos Until Gaia J. Andersen 1,2, B. Nordström 1,2 1 Dark Cosmology Centre, The Niels Bohr Institute, University of Copenhagen, Denmark 2 Stellar Astrophysics Centre, Aarhus University, Denmark
The Milky Way: Our Prototype Disk Galaxy Us NGC 1232 NGC 4565
Witnesses of the Evolution: Stars Detailed chemistry and motions of stars of all ages Disk probes: Solar-type dwarfs; Halo: Giants What parameters do we want to know? - Age; dwarfs (models, Teff, log g, Mv uvbyβ) - Metallicity ([M/H] ~ [Fe/H]; [X/Fe]) - Distances, reddening π; uvbyβ - Space velocities U,V,W μ, π, RV - Disk heating; Galactic orbits; inner/outer halo Main targets: Solar vicinity (disk life); extreme early halo
Rewinding History: The GCS Disk Survey I: uvbyβ photometry of all ~30,000 A5-G stars in the HD to V 8.5 Apparent magnitude limited sample (E. H. Olsen 1975-1994) II: Select all ~17,000 bona fide F & G dwarfs from stage I Ages can be derived for stars 0.5-1.5 mag above ZAMS When distances known, volume limited sample can be defined III: Multiple RV observations of (nearly) all stars from stage II (~63,000 new obs. of 14,139 stars - ~1,000 observing nights @ ESO + OHP + CfA; Geneva-Copenhagen team 1981 1997) IV: Re-check and complete all data (Hipparcos!); revisit calibrations (Re)compute distances, velocities, [M/H], ages, Galactic orbits (Lund Copenhagen Geneva 2003 2009)
The GCS Disk Sample; Completeness Volume complete to ~ 40 pc R ~ 100 pc Solar orbit Ill.: Lars L. Christensen, ESO
Preparation: Test the Stellar Models Key questions for ages at start: - Opacities; core convection; overshooting Test objects 1: Detached double-lined eclipsing binaries - Complete, accurate, uvby light curves - Spectroscopic orbits (1.52m spectra, CORAVEL, + ) - [Fe/H] ([M/H] from uvbyβ; [X/Fe] from CES, McD, ) Test objects 2: Open clusters with ages 1-4 Gyr - Accurate uvby colour-magnitude diagrams - RV identification of non-members and (binary) members
Method for Calculating Ages and Errors [Fe/H] Adjust Teff scale as needed to match observed ZAMS @ low [Fe/H]
MDF and Age-Metallicity Relation G dwarf problem and closed-box chemical evolution
AMR: Evolution Over 20 Years Edvardsson et al. (1993) Gaia-ESO Survey (2014) - Same selection effects
The U-V Plane: Not Two Gaussians! Hyades Not thin + thick disk! Dynamical focusing? Radial migration?
Disk Heating: Models vs. Reality Young Old Thick obs lin. Q&G Merger @ 3 Gyr Quillen & Garnett 2001: 189 dwarfs from Edv 93 GCS III: 2,626 single stars, σ п < 13%; σ Age < 25%
GCS I-III Finally Published 2004-2009 Set a deadline! Now ~1,600 cit. total
The GCS Movie Solar neighbourhood??
A Subset: Detailed Element Abundances What stars made which elements, how, where, and when? - Test the best atmospheres & synthetic spectra - Instrument (CES) designed to obtain accurate spectra First visitor run December 1981 - a young David Lambert
Disk: Evolution of Element Abundances Subsample of 189 FG dwarfs Edvardsson et al. (1993) Now 1,650 cit
General: Galactic Chemical Evolution From the Big Bang via the Solar System to Today Conventional picture: Pop. III stars were presumably massive, exploded as SN II, and efficiently returned new heavy elements to the ISM. This enabled lower-mass extremely metal-poor (EMP) halo stars with [Fe/H] ~ 3 to form in GMCs seeded with these elements). Later SNe II and eventually SNe Ia further boosted the chemical evolution of the Milky Way halo and disk at steadily higher [Fe/H]. Outliers were just due to binaries. Observational check: Precise ([X/Fe]) ratios of normal (Pop I +) Pop II stars ( Edv+ for the halo )
Stochastic Early Halo Enrichment? Conventional Wisdom: Stochastic chemical evolution models produce significant dispersion in EMP element ratios (e.g. [Mg/Fe]). Observational Reality: Precise observed [Mg/Fe] relations are tighter than models predict. Very accurate observations are needed: UVES@VLT!
First Stars : The Normal Pop. II Pattern Giants Cr II in dwarfs Element pattern O Fe group extremely uniform, but NLTE, 3D convection
The First Elements Produced: CNO C from CH band @ 430 nm N from NH band @ 336 nm Why is the scatter in C and N so enormous?? SNe II should be simple! Is this intrinsic or caused by internal processing?
Origin of the scatter in C & N? Mixed Unmixed O behaves much better? C/N: mixing with CNO cycle!
Explanation: CNO Cycle + Mixing! Li confirms the deep mixing - and the 12 C/ 13 C ratio as well CNO cycle just turned C into N Gaia π!
What True Chemical Peculiarities are Prominent in Today s EMP Stars? C-enhanced EMP (CEMP) stars: [C/Fe] = 0.7 2+ dex; 20-70% of EMP giants May - or may not - also show enhanced [r,s/fe] R-process enhanced stars (EMP-r stars): [r/fe] = 0.3 ~2 dex; rare: ~3% of EMP giants Some stars also have [r/fe] < 0 (e.g. HD122563) NB: Their spectra are complex, crowded, and non-standard!
Key: Local or Global Enrichment? Basically just two explanations of these excesses: This is only a surface effect, produced locally; or The parent ISM cloud had this composition Local production site(s): Production, diffusion or mixing within the star itself, or Transfer of processed matter from a binary companion External production sites: Pollution of the parent cloud from a distant production site Clue: Frequency and orbits of binaries in sample!
What Was Thought Before 2006? Conventional wisdom: All chemical peculiarities were ascribed to binary evolution Origin of EMP-r stars: Original primary star became a SN II; polluted companion Origin of CEMP stars: Original primary star became an AGB star; polluted companion Precise observational clues needed: Reliable binary frequencies; orbital periods & eccentricities
Paradigm: C and Ba+s Likely Transferred From a Former AGB Binary Companion Pop II (Jorissen et al. 1998) Note cutoff periods for: common-envelope evolution (secondary R) circularisation of orbits (Pop I: Depends on age) Theoreticians happy
Do EMP-r (r-ii) Stars Have a Similar Origin? SS r-pattern SS s- pattern [Fe/H] = 3.0±0.2; r-element pattern very uniform, but offset by ~2 dex
RV Monitoring With FIES@NOT (I) Both the prototype r-ii+cemp star CS22892-052 and the C normal U star CS31082-001 are single (σ 100 m s -1 ) rare elements are irrelevant(!) One U star has K ~ 350 m s -1 One r-i star: P ~7 yr; e ~ 0.77 P ~ 300 d
CEMP Stars Were Now Important: CEMP Stars Come in Two Varieties: CEMP-no Stars: No s-element signatures Most metal-poor group ([Fe/H] 3) Dominant group in outer halo Binary properties unknown CEMP-s stars: Strong s-element signatures Dominate in inner halo Binary frequency high (simulations said 100%??)
RV Monitoring With FIES@NOT (II) Most ( 80%) CEMP-s stars are indeed binaries, but CEMP-no and EMP-r stars are generally single, and some (~20%) CEMP-s stars are single as well (σ < 100 m s -1 ). P = 20 d (post common envelope?) P ~30 yr(!)
Follow-up I (2017+): Gaia π, d, L Distances Kinematics Evolution - are unknown! Are the stars Giants or AGB? Some CEMP-s stars are variable Answer in 2017: Gaia space astrometry & photometry!
Tantalizing Results From the LMC: Pulsating LMC RGB & AGB stars from MACHO and Spitzer (Riebel+ 2010). Sharp limit between RGB and AGB stars? Gaia parallaxes will help to put our field stars on the same M v scale.
Results/Conclusions: CEMP-no and EMP-r stars are basically single ~20% of the CEMP-s stars are single as well(!) Most, but not all CEMP-s stars are in (long-period) binaries Abundance anomalies are intrinsic and were imprinted on the parent clouds across interstellar space in ISM at z 3(?) Some early enrichment processes were complex and non-local Could this process account for the C-rich DLAs at z = 2-3? Could the C in CEMP-no and the single CEMP-s stars originate in AGB stars without s-element production? Alternatives: Faint SNe with fallback & mixing? Or spinstars? Some CEMP stars seem to be pulsating; Gaia should tell if & why!
THANK YOU!